The TF3MA 20m to 10m band dipole antenna

I wanted to make a simple wire dipole antenna which covered all ham bands from 20 to 10 meters and could be hung up in a tree and fed with 50 Ohm coax without using 4:1 Balun (but 1:1 current balun OK).

The idea

Using NEC software I analyzed the classic multi-band wire dipole and found that the bandwidth on each band is limited and the SWR is very high outside the bands, so I decided to try different design.
Based on the ideas from the ZL-special antenna (2-el log periodic beam) and several LPDA antennas, I wanted a SWR like the of a LPDA but with fewer wire elements (maximum 3), plus I wanted a dipole type radiation pattern on all bands, (no beam like pattern you get using multi band doublet or lazyH).

NEC modeling

I started the modeling with 2 wire dipole elements separated by 0.4 meters and connected with crossed 300 Ohm twinled, like the ZL-special, but the dipoles cut for 20 and 15m bands.
This construction covers both bands with low SWR when fed from 50 Ohm source, gain is approx 6.5 dBi. Looks good....
I then added the third dipole for 10m band and it hand to be placed 0.6 meters from the 15m dipole to get low SWR on 10m band. Now I have an antenna working on all 3 bands with low SWR and 50 Ohm feed point impedance.
Think of this design as a ordinary log periodic dipole array, but throw away all but the 3 dipoles covering the 20, 15 and 10 meter bands.

Construction and testing

I cut the dipole wires and connected to the 300 Ohm twinled (better if you can use strong 450 Ohm twinlead, I have had problems with the small gauge 300 Ohm line) and coax and tied mono-filament line between the wires for proper spacing of the wires. I then hung this wire structure between two 9m high trees and proceeded to measure the SWR. The SWR curve turned out as expected, but I had to trim the lengths to get the lowest SWR centered on the ham bands. Had some QSO's using it on 15m band during the 2002 SSB ARRL DX contest.

Results

The SWR is low (<2, typ <1.5) on 20, 15 and 10m bands, and the SWR is low over the whole band (20, 15, 10), unlike traditional multiband dipole antennas, which cover only a part of each band with low SWR.
Here are the measured SWR values for the prototype antenna. The SWR is at the transceiver end (approx 20m RG58C/U cable) Yes, the dipoles need be shorted a little more to move the lowest SWR towards the middle of each band.

 F    SWR     F    SWR     F    SWR     F    SWR
14.0 1.30    21.0 1.15    28.0 1.07    28.8 1.80
14.1 1.37    21.1 1.18    28.2 1.13    29.0 1.90
14.2 1.50    21.2 1.25    28.4 1.25
14.3 1.60    21.3 1.30    28.6 1.50

AND I can use the antenna with tuner on the 17m and 12m band as the SWR on those bands is on the order of 3, the 20m length of RG58C/U coax I use has attenuation of 1 to 2 dB and this lowers the SWR observerd at the feeding end.

Wires and balun

The wire I used for this antenna is 1.6mm dia insulated multi strand wire. The 20 meter antenna can be constructed using selfsupporing 50mm or less dia aluminium tubing with the rest of the antenna using the 1.6mm dia wire. The lengths listed are approx lengths and have to be trimmed for each band using a SWR bridge. The antenna is fed via current balun made of 8 turns of RG58C/U on 10cm dia form. On top is the 20m dipole and the 10m dipole is lowest

Making a strong 300 Ohm twinlead using 4mm dia wire.

I found a simple way of making the 300 Ohm twinlead using a piece of 4mm dia multistrand insulated wire.
Make 3 spacers (10mm*10mm) out of soft plastic (I used IKEA cutting boards), each about 100mm long and 6 spacers abt 80mm long, Drill 4mm dia holes for the dipole wire about 15 mm from the end half way through the insulators and also holes from the ends into these so that the dipole wires can be thread through the ends. Drill two ~6mm dia holes (close fit for the TL wire) with 50mm spacing. I used abt 100mm between spacers.
On each wire I put soldered terminal ring connectors and connected the wires to the TL using 5 mm rostfrei hardware.
Remember to cross the wires from the middle dipole. Then insulate the connections with vulcanizing rubber tape. This makes a strong semi-flexible 300 Ohm feedline.
The current choke I made by winding 8 turns of RG-58 on a 100mm dia form which I secured to the support/spacer for the 10m dipole.

300 ohm TL and 20m dipole connection

300 ohm TL and Balun and 10m dipole connection

The dipole drawing

NEC2 Simulation results

Nec2 input file for simulation

CM NEC Input File 
CM TF3MA multi-band Dipole 
CE
GW  1  21    0.00000   4.90000   9.00000   0.00000   -4.90000  9.00    0.0008
GW  2  17    0.00000   3.15000   8.60000   0.00000   -3.15000  8.60    0.0008
GW  3  11    0.00000   2.50000   8.00000   0.00000   -2.50000  8.00    0.0008
GE  1
EK 
GN  0  0    0    0  13.00E+00  5.00E-03  4.30E+00  1.00E-03  0.00E+00  0.00E+00
FR  0  80  0    0  10.00E+00  0.25E+00  0.00E+00  0.00E+00  0.00E+00  0.00E+00
LD  5  1    0    0 3.720E+07
LD  5  2    0    0 3.720E+07
LD  5  3    0    0 3.720E+07
TL  1  11   2   9 -300
TL  2  9   3   6 -300
EX  0  3    6    0  1.00E+00  0.00E+00  0.00E+00  0.00E+00  0.00E+00  0.00E+00
RP 0, 37, 72, 1000, 0, 0, 5, 5
PT  -1
EN

Xnecview output of simulation

Radiation pattern at 14.0MHz vs height, and vs vertical dipole

Radiation pattern at 14.0MHz for verical dipole and LPDA for heights of 6m, 8m, 10m, 12m and 14m. For best results put the antenna up to at least 10m height.
For comparision i inserted the radiation pattern for a vertical dipole (the innerst curve), with the lower end 1m from ground, it has a gain about 0dbi with normal ground parameters (eps=13.0 and sigma=0.005).

UPDATE - Oct 11th 2004 --- Actual wire dimensions

I made the LPDA out of heavier wire for strength. The actual dimensions for this version are:

20m dipole --- 4.70 m per leg of 2.5mm2 insulated multistrand wire
               |
               TL -- 0.45m 4.0mm2 insulated multistrand wire -- spaced by 50mm
	       |
15 dipole --- 3.10 m per leg of 2.5mm2 insulated multistrand wire -- crossed!!!
	       |
               TL -- 0.40m 4.0mm2 insulated multistrand wire -- spaced by 50mm
	       |
10m dipole --- 2.45 per leg m of 2.5mm2 insulated multistrand wire
	       |
           Coax balun 8 turns of RG-58C on 100mm dia form.

This is a stronger design but the SWR curve matches the NEC 2 simulation quite well. The measured SWR at the TX/RX using abt 15m of RG-58c coax from the LPDA is as follows (using MFJ-269):

  SWR    Freq
  2.0 --- 13.6 MHz
  1.0 --- 14.0 MHz
  2.0 --- 15.0 MHz
  3.3 --- 18.1 MHz
  2.0 --- 20.0 MHz
  1.3 --- 21.3 MHz
  2.0 --- 22.5 MHz
  3.6 --- 24.9 MHz
  2.0 --- 27.0 MHz
  1.2 --- 28.3 MHz
  2.0 --- 29.4 MHz

Here is the calculated SWR and gain

and the NEC2 model

CM NEC Input File for tf3ma lpda
CM Dipole 
CE
GW  1   12    0.00000   4.70000   9.00000  -0.02500   0.00000  9.00    0.002
GW  11  12    0.00000  -4.70000   9.00000   0.02500   0.00000  9.00    0.002
GW  2   8     0.00000   3.10000   8.55000   0.02500   0.00000  8.55    0.002
GW  21  8     0.00000  -3.10000   8.55000  -0.02500  -0.00000  8.55    0.002
GW  5   1     0.02500   0.00000   9.00000   0.02500  -0.00000  8.55    0.003
GW  51  1    -0.02500   0.00000   9.00000  -0.02500   0.00000  8.55    0.003
GW  3   6     0.00000   2.45000   8.15000  -0.02500   0.00000  8.15    0.002
GW  31  6     0.00000  -2.45000   8.15000   0.02500  -0.00000  8.15    0.002
GW  6   1    -0.02500   0.00000   8.55000  -0.02500  -0.00000  8.15    0.003
GW  61  1     0.02500  -0.00000   8.55000   0.02500   0.00000  8.15    0.003
GW  7   1    -0.02500   0.00000   8.15000   0.02500   0.00000  8.15    0.002
GE  1
EK 
GN  0  0    0    0 13.00E+00  5.00E-03  4.30E+00  1.00E-03  0.00E+00  0.00E+00
FR  0  72   0    0 12.00E+00  0.25E+00  0.00E+00  0.00E+00  0.00E+00  0.00E+00
LD  5  1    0    0 3.720E+07
LD  5  11   0    0 3.720E+07
LD  5  2    0    0 3.720E+07
LD  5  21   0    0 3.720E+07
LD  5  3    0    0 3.720E+07
LD  5  31   0    0 3.720E+07
LD  5  5    0    0 3.720E+07
LD  5  51   0    0 3.720E+07
LD  5  6    0    0 3.720E+07
LD  5  61   0    0 3.720E+07
LD  5  7    0    0 3.720E+07
EX  0  7    1    0  1.00E+00   0.00E+00 0.00E+00  0.00E+00  0.00E+00  0.00E+00
RP  0  72  72 1001 -90.00E+00 90.00E+00 5.00E+00  5.00E+00  0.00E+00  0.00E+00
PT  -1
EN